Discover Impacts with AI

Instead of recording every climate impact manually or via a checklist, you can have an AI provider suggest them for the current climate DMA assessment. The AI evaluates your organisation profile; you review each suggestion editorially and decide which impacts are actually created.

Prerequisites

• An AI provider is configured under “Settings > Integrations”. Without a configured AI integration, the menu item “Discover with AI” does not appear at all.
• The organisation profile is complete – in particular, a “NACE code” is stored.
• The climate DMA assessment is in Draft status and you have editing permission.

Step by step

1. Open the “Climate DMA” page and scroll to the “Climate impacts” section.
2. Open the “Add impacts” dropdown button and select “Discover with AI”. The “AI-suggested climate impacts” dialog opens on the right.
3. In the dialog, click “Discover with AI” to start generating suggestions. A loading animation appears during generation.
4. If the organisation profile is incomplete, the application stops and lists the missing prerequisites, for example Organisation profile - NACE code missing. Add the missing master data and restart the process via “Try again”.
5. If generation is successful, a list of editable suggestions appears across the three ESRS E1 subtopics – positive and negative, actual and potential impacts.
6. Review each suggestion individually: use “Accept” to adopt it, or “Change” to adjust it before adoption (description, classification, severity values). Suggestions you do not accept simply stay unedited – they are not created.
7. At the bottom right, click “Add N accepted impacts”. The button becomes active as soon as at least one suggestion has been accepted, and creates the accepted suggestions as regular climate impacts.
8. Use “Cancel” to close the dialog without adopting any impacts.

What happens next

• Adopted suggestions appear as regular entries in the “Climate impacts” table and cannot be distinguished from manually recorded impacts – you can assess and edit them as usual afterwards.
• The AI call is recorded in the internal AI interaction log, and the process is additionally logged in the activity history.

Practical tip: Before the first AI run, maintain the organisation's business description and NACE code. The more precise the profile, the more industry-specific the suggested impacts will be – this saves rework when sorting out generic suggestions.